Building a Hyperconverged Campus IT Infrastructure
The next generation of IT infrastructure, hyperconvergence combines computing, storage and networking into a single, simplified and automated system that is far easier and less costly to use and maintain. It is a perfect fit for colleges and universities, which often have limited IT resources but have enterprise-level IT needs.
This web seminar explored hyperconverged IT infrastructures, described how these systems work in higher ed environments, and outlined the keys to successful implementation and deployment. Hyperconverged IT infrastructures can benefit institutions by dramatically reducing IT costs and complexity, simplifying app delivery and reducing the workload on IT staff.
Mike Lyon: About 15 years ago, we stopped buying physical servers, and virtualization started coming into the scene. We did that to simplify our lives. We didn’t want to have to manage an ever-evolving and ever-growing server farm. It was a waste of resources. It was getting costly, and the data center costs for having to physically add additional servers became pretty tedious.
So we wanted to simplify that, and in a way, we did. We didn’t have to manage the 10x factor of physical infrastructure that we had before. But instead of managing a farm of physical servers, we were managing these complex and expensive SANS with their own set of expertise that was required. Then we were also managing the beefier physical hosts. On top of that, we had a separate layer for the hypervisor, which introduced a new layer of management outside of managing storage and outside of managing physical hosts.
So in a way, we simplified our lives, but we just changed where we were spending our time. Instead of doing it in thousands of servers, we just did it in fewer, more complex and expensive components. So what would be the next evolution, and how could we simplify that?
Think about what the cloud has done and about why everyone was so quick to adopt its simplicity. The ability to deploy an X86 server in a few clicks is amazing because you don’t have to worry about what’s on the back end and where that storage is coming. That cloud-like simplicity is what hyperconverged infrastructure is aiming for, but with on-premise hardware, it’s in your physical control and in your data center.
The definition of hyperconverged infrastructure (HCI) is combining the server storage and hypervisor into one solution architecture. They all interact with each other natively without having to use an external SAN or an external VM to manage the storage. It’s all these layers in one architecture. It’s much like you get in a public cloud setting, but it’s a private cloud. Simplicity, scalability, availability and affordability are the four tenets of what a successful HCI implementation looks like.
Single vendor and a unified support path also come up with customers. They love not having to deal with multiple vendors when there’s a problem. A hyperconverged stack owner has the responsibility of providing world-class support because they own the entire stack. They cannot offshore the burden to another infrastructure component. Also, it should make the integration less than half of what you’re used to. You’re getting rid of the infrastructure expertise.
The pace at which technology changes is always accelerating. Students have expectations that every piece of technology will natively work. How do we ensure we have a platform that can accommodate today’s growth and tomorrow’s growth? There isn’t a crystal ball to forecast the technology needs even in 12 months, so it goes back to having the ability to scale and grow as you need without making your life more complicated. You want to preserve simplicity as you grow, and keep the cost of care and feeding of the infrastructure less than or equal to what they were before.
There are a lot of advantages. Keep the support path the same. Have one interface to manage your entire infrastructure. There’s simplicity in management—having one interface to view and control. And then there’s building in disaster recovery.
In a survey of our users about the advantages they see, 90 percent reported a decrease in time spent recovering from a failure. Eighty percent talked about DR and how that was able to give them peace of mind. And 100 percent said that having single-vendor support was crucial. It’s important that you have the confidence to be able to call your vendor and not worry about getting pushed off onto someone else. That’s paramount to having a good HCI experience— from setup and deployment and implementation to an initial support call or whatever type of troubleshooting you may need to be doing.
To watch this web seminar in its entirety, please visit universitybusiness.com/ws072418